Trey-Less Efficiency Explains Why Duke is Perceived Class of Final 4
Fantastic Fundamental Play Through Dance with Healthy Cooper Flagg
I’m having a bear of a time trying to estimate Power Ratings for these last four college basketball teams in the NCAA Tournament. It’s usually pretty easy. This year, Duke is currently priced head-and-shoulders above Houston, Florida, and Auburn even though most respected rankings wouldn’t have them far apart.
Duke is currently -5 vs. Houston in Saturday’s semifinal matchup
Kenpom has Duke winning by 2 points
Yes, the public loves betting Duke…and sports books have to guard against public money. But, there’s usually not THAT big a difference between quant projections and how the public bets. It’s very likely that there are sharp methodologies that show Duke -2, or -3, or even -4 is a smart bet or the line wouldn’t be that high. At the very least…if Duke -5 was “too high,” sports books would have been POUNDED with a lot of Houston money already. “I can get the Cougars +5 when most models make it below that…I gotta bet that RIGHT NOW!”
No money that matters is betting Houston yet. That suggests Kenpom is too low…and that there must be myriad methodologies suggesting his projection is too low. I’ll talk more about Power Ratings later in the week. For now, let’s try to figure out why Duke is getting so much respect.
So far through the NCAA tournament, Duke is playing at a VERY high level. That’s easily expressed with the Trey-less efficiency stat we’ve been using in recent weeks (which I mentally invented awhile back sitting through a 45-minute speech in Mandarin). It’s simply what happens in a basketball game on the possessions that DON’T end with a made three-pointer. Manages to simplify the impact of two point offense and defense, rebounding, turnovers, and the ability to earn free throw attempts into one quick an easy number. It also manages to factor in the penalty of MISSING too many three-pointers. Imagine an old-school stereo console that used to have five or six knobs you had to turn to get the music right…but now has just one. Trey-less efficiency has proved to be a very handy dial.
Let’s look at Duke’s Trey-less efficiency marks by game through the postseason. We’ll do that for all Final Four teams in this article. Tricky thing here is that Duke was missing its star player in the ACC tournament. You’ll see the impact of that right away. I’ll also put made three-pointers in parenthesis for each team and their opponent so we can get a sense of THAT dynamic as we run through today’s numbers. Oh, I’m going to leave OUT all the Dance openers against #16 seeds because those were blowouts vs. irrelevant non-entities that might accidentally muddy the picture.
DUKE
Duke .95, Georgia Tech .62 (7-12) (Flagg 15 minutes)
Duke .94, North Carolina .95 (5-3) (Flagg out)
Duke .80, Louisville .72 (10-7) (Flagg out)
Duke 1.13, Baylor .78 (12-8)
Duke 1.12, Arizona .97 (11-12)
Duke 1.00, Alabama .63 (6-8)
3-game average w/Flagg: Duke 1.08, Opponent 79.3 (9.7 to 9.3 treys)
Big jump in offensive explosiveness once Flagg returned. Defense was pretty darned good too outside of the Arizona game. It’s not great to have only a three-game sample size for evaluation. But…those three games are telling the same approximate story. Duke is super-effective inside the arc, while ALSO being capable of making 11-12 treys in a big game. JUGGERNAUT! Let’s run through the other teams so we have some reference points. I’ll use conference tournament games and Dance action from the second round on for everybody.
HOUSTON
Houston .88, Colorado .84 (8-5)
Houston .83, BYU .63 (10-6)
Houston .79, Arizona .80 (9-5)
Houston 1.00, Gonzaga .88 (8-9)
Houston .73, Purdue .73 (9-8)
Houston .84, Tennessee .65 (9-5)
6-game Average: Houston 84.5, Opponent 75.5 (8.8 to 6.3 treys)
Not nearly the explosiveness of Duke. In fact, VERY CONSISTENTLY not the same as Duke offensively outside of the Gonzaga outlier. Fantastic defense though. And, better trey differential that seems keyed by superior defense around the arc. Since Houston and Duke are playing each other, let me stack those…
3-game average w/Flagg: Duke 1.08, Opponent 79.3 (9.7 to 9.3 treys)
6-game Average: Houston 84.5, Opponent 75.5 (8.8 to 6.3 treys)
Two top teams…but Duke’s differential is much better (call it +29 to +9 by throwing out the decimals). Houston’s capable of making up for that if it truly has a defense that can keep Duke from making treys. But, you can also say that Houston seems much more mortal without a trey advantage (clear in the Purdue game), while Duke can still crush people if it doesn’t have a trey edge.
Both Houston and Duke played Arizona on a neutral court. Let me stack those for comparison…
Duke 1.12, Arizona .97 (11-12) (Duke won 100-93)
Houston .79, Arizona .80 (9-5) (Houston won 72-64)
Different types of games with different styles. But, you can see that Houston had to win treys by 12 points to get scoreboard distance. Duke won its Dance encounter with Arizona by seven points despite losing treys by three points. Blue Devils more impressive at the “other stuff” besides treys in this common opponent comparison.
Dance-quality Baylor can also go in as a common opponent. Duke beat the Bears badly (89-66). Houston played them twice during the regular season (one home, one away). Let me merge those into one game to approximate a neutral court equivalent.
Duke 1.13, Baylor .78 (12-8)
Houston .89, Baylor .74 (18-18 in two games, so 9 per game each)
Duke’s defense shows up as comparable to Houston’s in this comparison. Much more explosive offense. Duke’s not going to be THAT good every time out. But, Duke’s 89-66 win was much more impressive than Houston’s victories of 76-65 and 65-61.
Much easier to see now why Duke is a 5-point favorite! Blue Devils really are thriving at the stuff that matters…and are doing so in more impressive fashion that Houston. Houston’s great. Not hard to make a case that Duke is about five points better in methodologies that emphasize recent form. Kenpom seems to be understating that.